Santa, Where Do Best Practices Come From?

Much as we like to believe that our compensation processes are objective, even numbers have a way of being influenced by human nature. If we are not aware of them, our vulnerabilities can affect our decisions all the way from job pricings to cost projections to talent assessments. After all, much of what we deal in is probability, and other forms of statistics, and we all know what kind of reputation statistical interpretations have.

One proven foible of our quantitative work is worth being talked about openly and often. We're likely to pay more attention to findings that support our beliefs and to ignore competing facts and patterns in our data. This is called the "confirmation bias," and we all suffer from it, with personality traits and other social influences impacting our bias.

Think about the wrangles you have with certain executives about salaries, from their own to their teams'. The exec is often waiting for you to confirm his/her beliefs, and has little interest in competing facts. We suffer from the same habits, we just exhibit them  less loudly.

Even though our own personal confirmation bias is not that obvious, it's there whether we like it or not. We often tell our bosses or internal clients how we expect our work to turn out. (Tell the truth: We do the same within ourselves with our inner voice, whether we are aware of it or not.) Of course, our predictions are based on experience but they can just as easily be incorrect.

Notice how high we set the bar when we run into evidence against our own hypotheses. This tendency is called the "disconfirmation bias" and it has also been supported through experiments.

You may think these biases have little importance to your work, but you can literally make mistakes that cost money and cause mischief if you do not formally examine your thought process each time you are asked to give a prediction. Fortunately, there are time-tested actions that you can follow including these three listed in a recent Harvard Business Review "Management Tip of the Day."

Start by following a data and analytic approach that is specified ahead of time. Document how this approach is followed -- including how judgments should be made using the data -- and hold everyone who is doing data analysis on the project to follow the same approach. This will govern the tendency we all have to cherry-pick data that agrees with our point of view and, for the same reason, the full process should be itemized for the person to whom you present the data.

Another step to include in your approach is to actively look for findings that disprove your belief. This is often called "shooting holes in the data," where you identify other patterns in the data and determine whether they indicate that other judgments could be more correct, or correct in a different way, from your own.

Treat your findings like predictions and test them. Even a single job pricing can benefit from this validation process -- how high can it go and how fast? What does this say about pay for other jobs that share the same salary grade? And so on.

In my experience, in-house compensation staff have the hardest time with the last two actions on the list. In comparison, while many in our field are biased against consultants, you can count on most to challenge potential biases. In fact, that's part of the rationale for hiring a consultant. We're trained to do it, and consider our bias checks as part of the service that a client is paying for. (Not all clients like this part of our service, by the way.)

For more on this particular topic, check out the recent WorldatWork video, "Advice on Working with Consultants from Consultants." And for the most jaded consultant-basher -- you know who you are -- allow the speakers to challenge your disconfirmation bias!

Margaret O'Hanlon, CCP collaborated with Ann Bares and Dan Walter to create the DIY guide to compensation leadership, Everything You Do (in Compensation) Is Communications @ https://gumroad.com/l/everythingiscommunication. Margaret is founder and Principal of re:Think Consulting. She brings deep expertise in compensation, communications and leadership to topics like the CEO Pay Ratio, performance management and compensation implementation discussions at the Café. Margaret is a Board member of the Bay Area Compensation Association (BACA). Before founding re:Think Consulting, she was a Principal at Willis Towers Watson. 

 

 

Comments

[jay]

Margaret, this is so good I can't pass it up. You develop the most wonderful things. I know I am 'testing the limits' of acceptance here but I just must comment on this good work.

The current field of compensation (total rewards?) is based NOT on 'best practice' in terms of what makes organizations thrive, but "Me to practices" that follow what others do without consideration of whether it fits your organization's business and talent needs at all. I guess we believe that if we 'dress like Marilyn Monroe' we will BE "Marilyn Monroe".

A few years ago we gathered considerable data from structured interviews with CEOs (published in Worldatwork Journal). These leaders wanted human resource to deliver reward practices that helped them realize the role people played in their specific organizations in achieving goals and subsequently sharing in the results of this success.

While our profession is focused only on copying what others do, our leaders (verified by hard data) want us to think like business people and become a member of the leadership team rather than behaving like what one CEO called, "Preschool Teachers" telling us the 'bogyman' is waiting to get us.

So until we redefine 'best practices' in terms that help businesses achieve results most of the 'copy-cat' surveying is not serving our organizations well.

You are a wise advisor, Margaret. I will buy you a gin and tonic some day.

(You will all be saved from my ranting until after January 1, Pat and I are traveling for our Xmas Cheer this year)